Transvaginal mesh implants independent review: final report

Chapter 4: Assessing the safety and effectiveness of vaginal mesh surgery for stress urinary incontinence and pelvic organ prolapse in Scotland, using nationally available NHS data

Update Since Interim Report

The Information Services Division ( ISD), part of NHSScotland, undertook an analysis of nationally available hospital discharge records on behalf of the Scottish Government's Independent Review of Transvaginal Mesh Implants over 2015 and 2016.

The analysis examined how many women undergo operations for SUI and POP, and how many go on to develop immediate or later complications, or require further surgery.

This chapter provides a summary of the final results. Full details of methods and results are available in the Lancet, as published on 20 December 2016:

http://www.isdscotland.org/whats-new/Stress-Urinary-Incontinence-and-Pelvic-Organ-Prolapse/

Preliminary results of this analysis were included in the interim report of the IR, published in October 2015. These final results differ from those published in the Interim Report, the main reasons being:

• The study period has been extended from 1997-98-2013-14 to 1997-98-2015-16.
• Enterocele procedures ( OPCS4 code P23.4) are now counted in the posterior (rather than anterior) non-mesh colporrhaphy group.
• Vaginal vault prolapse repair procedures involving transvaginal placement of mesh ( OPCS4 code P24.6) are now labelled 'Vaginal mesh vault repair' (rather than 'Infracoccygeal colpopexy').
• Further incontinence and prolapse surgery occurring after the index procedure are now considered as two separate outcomes (rather than a single composite outcome).
• The secondary outcomes initially examined (for example all readmissions, referrals to pain clinics, and prescriptions for pain relief) added little to the main outcomes (complications and further surgery) so have not been included in the final analyses.
• Analysis errors have been corrected, in particular:
  • Previously procedures were excluded if the woman had undergone prior incontinence or prolapse surgery in the five years up to 1997/98. Now they are excluded if the woman had undergone prior incontinence or prolapse surgery in the five years up to the procedure being considered. This means that repeat procedures are now more effectively excluded from the analysis as intended.
  • Previously vault prolapse repair procedures were excluded if the woman's prior surgery had involved any incontinence or prolapse procedure. Now vault procedures are excluded if the woman's prior surgery involved any incontinence or prolapse procedure except hysterectomy. In addition, vault procedures were previously included regardless of any other incontinence or prolapse procedures done at the same time. Now vault procedures are only included if done as a single procedure or in combination with a standard (non-mesh) anterior and/or posterior colporrhaphy. This means that 'first, single' vault repair procedures are now included as intended.
  • Previously only a certain subtype of immediate complications (procedure related) was counted in the total number of immediate complications. Now all subtypes (haemorrhage, infection, pain, procedure related) are included as intended. This means that our results now show a higher proportion of women experiencing immediate complications.

4.1 Operations provided in Scotland for stress urinary incontinence and pelvic organ prolapse

For this study, ISD used routine hospital discharge records to identify the different operations provided for SUI and POP in Scotland between 1997/98 and 2015/16. Specific types of operation that were provided in reasonably high numbers were included in the analysis.

In general, only single operations were included in the analysis. 'Single' means that the woman did not have any additional/second operation for incontinence or prolapse at the same time as the operation being examined. It is quite common for women to have more than one operation at the same time. However, if complications subsequently develop it can be difficult to know which operation caused the problem. Only single operations were included so that the study could focus on the risks of each particular operation separately.

In general, only first operations were included in the analysis. 'First' means that the woman had not had any other operation for incontinence or prolapse in the previous five years. In addition, only the first potentially eligible operation provided within the study period was included for any individual woman. Only first operations were included because the risk of complications may be quite different for a woman having a repeat operation. It was important that the study did not mix operations with different levels of risk.

4.2 Operations provided for stress urinary incontinence

In the late 1990s open colposuspension was the main operation provided in Scotland for SUI (around 500 first single procedures per year). Tape (mesh) procedures were introduced in the UK from 1998 and quickly replaced colposuspension as the most common operation type for this condition. However, the number of tape procedures fell substantially in the most recent years included in the analysis. First single urethral injection therapy and suprapubic sling operations have been provided in moderate numbers (fewer than 100 per year) throughout the time period included in the analysis.

Operations provided for stress urinary incontinence

Operations provided during a patient's admission to hospital are recorded on routine hospital discharge records using OPCS Classification of Interventions and Procedures codes. Between 1997- 98 and 2005-06, the codes available did not specify which kind of mesh tape operation had been provided. After April 2006, new codes allowed the particular type of mesh tape operation (retropubic or transobturator) to be recorded.

4.3 Operations provided for pelvic organ prolapse

Anterior and posterior non-mesh colporrhaphies (first, single operations) have been commonly provided (up to around 500 per year) throughout the study period. Anterior and posterior mesh colporrhaphies can be identified in hospital discharge records from 2007/08 onwards. Relatively small numbers of mesh colporrhaphies have been provided in Scotland since then, and numbers provided have fallen in the most recent years.

Sacrospinous fixation operations have increased markedly over recent years (to around 300 first single procedures per year). Vaginal mesh vault prolapse repair procedures can be identified in hospital discharge records from 2006/07 onwards. Relatively small numbers have been provided since then, and numbers provided have fallen in the most recent years. Mesh open sacrocolpopexies have been provided in moderate numbers (fewer than 100 per year) over the time period included in the analysis. Moderate numbers (around 100 per year) of vaginal hysterectomies for POP have also been provided over the time period included in the analysis.

Operations provided for pelvic organ prolapse

Sacrospinous fixation, vaginal mesh vault repair, and open sacrocolpopexy are provided for prolapse of the top of the vagina following a hysterectomy. These operations were therefore included if the woman had had a previous hysterectomy (but no other operation for incontinence or prolapse in the previous five years). In addition, these operations are rarely done as single operations so ISD included them if they were done at the same time as a traditional (non-mesh) colporrhaphy (but no other incontinence or prolapse operation).

Vaginal hysterectomy can be done for prolapse or other problems such as heavy periods. Only vaginal hysterectomies done for prolapse were included in the analysis.

4.4 Problems after surgery for stress urinary incontinence or pelvic organ prolapse

Main problems

ISD looked at four main categories of problem that can develop after an operation for SUI or POP. These were:

• immediate complications;
• later complications;
• further incontinence surgery; and
• further prolapse surgery.

Immediate complications

'Immediate complications' means that at least one complication was recorded on the same hospital discharge record as the operation being examined; in other words, the woman developed a complication when she was still in hospital following her first operation.

Later complications

'Later complications' means that at least one complication was recorded on a subsequent hospital discharge record; in other words, the woman had been discharged home then readmitted for a complication at a later date. In general, readmissions for later complications were counted if they happened within five years of the operation being examined. Complications that would be expected to develop quickly after an operation were only counted if the readmission was within three months of the operation.

Further incontinence or prolapse surgery

'Further incontinence or prolapse surgery' means that at least one operation for either of these conditions was recorded on a subsequent hospital discharge record; in other words, the woman had been discharged home after her first operation then readmitted for another SUI or POP operation at a later date. All readmissions for further surgery were counted if they happened within five years of the operation being examined.

What is a 'complication'?

'Complications' included the following:

• problems directly related to the operation, such as damage to the bladder or difficulty passing urine;
• excessive bleeding;
• infection;
• pain; and
• partial or total removal of mesh (later complications only).

Only complications that were treated in hospital were included in the analysis.

Complications treated in outpatient clinics or in general practice were not included because this information was not available to us.

The risk of developing problems after an operation

The risk of developing problems after an operation for SUI or POP depends on the type of operation done and on a number of other factors such as:

• the age of the woman;
• how many additional health problems she has; and
• how experienced the surgeon doing the operation is.

To compare the risks specifically associated with different types of operation, it is important to take account of these other factors that may be influencing the number of problems seen. For example, if older women with a lot of additional health problems tend to have mesh colporrhaphies rather than standard (non-mesh) colporrhaphies, we would expect to see more problems after mesh operations even if mesh colporrhaphy did not in itself carry any more risk than standard colporrhaphy.

Statistical methods can be used to take account of other factors that may influence the number of problems seen after different types of operation and allow us to focus on the differences that are due specifically to the type of operation that was provided.

4.5 Problems following operations for stress urinary incontinence

The risk of developing problems after the different types of SUI operation included in the analysis is shown below.

Problems following operations for stress urinary incontinence

This is the total number of readmissions that would occur on average if 200 women were each monitored for five years after having their SUI operation.

The increase or decrease in risk of the various problems following each type of operation compared to that experienced by women undergoing open colposuspension, the commonest non-mesh operation, is shown below.

These final results have used statistical methods to take account of various factors that may influence the level of problems seen after operations as discussed above. The factors that have been accounted for are women's age, deprivation level, and additional health problems; the experience of the surgeon; and the type of hospital providing the operation.

Taking these factors into account means that the remaining differences in risk are not due to those factors and are likely to reflect genuine differences in risk associated with the different types of operation.

To help interpret these figures, a 50% decrease in risk is the same as the risk being halved, and a 100% increase in risk is the same as the risk being doubled.

Change in risk of problems following specific type of operation compared to that experienced by women undergoing open colposuspension

Green indicates significantly lower risk than that seen after open colposuspension
Red indicates significantly higher risk than that seen after open colposuspension

4.6 Summary of findings for stress urinary incontinence operations

Mesh tape procedures for SUI carried a lower risk of immediate complications than open colposuspension (a non-mesh procedure). Infections and problems directly related to the operation were the most common immediate complications following all types of SUI operations.

Mesh tape procedures carried a similar risk of being readmitted for a later complication compared to open colposuspension.

The relatively high risk of later complications seen after urethral injection therapy (a non-mesh procedure) may be due to the very high risk of needing another incontinence operation after this type of surgery (see below).

Problems directly related to the operation, infections, and (for mesh operations) further surgery to remove the mesh, were the most common later complications seen after operations to treat SUI.

Mesh tape procedures carried a similar risk of being readmitted for repeat incontinence surgery compared to open colposuspension. By contrast, urethral injection therapy carried a much higher risk of being readmitted for further incontinence surgery than open colposuspension.

All the procedure types examined carried a lower risk of being readmitted for subsequent prolapse surgery compared to open colposuspension.

4.7 Problems following operations for pelvic organ prolapse

The risk of developing problems after the different types of POP operation included in the analysis is shown below.

Problems following operations for pelvic organ prolapse

This is the total number of readmissions that would occur on average if 200 women were each monitored for five years after having their POP operation.

The increase or decrease in risk of the various problems following each type of operation compared to that experienced by women undergoing anterior colporrhaphy (the commonest non-mesh operation) is shown below.

As described earlier, these final results have used statistical methods to take account of the various other factors that may influence the level of problems seen after these operations. The differences shown are likely to reflect genuine differences in risk associated with the different types of operation.

Change in risk of problems following specific type of operation compared to that experienced by women undergoing anterior colporrhaphy

Green indicates significantly lower risk than that seen after anterior colporrhaphy (non-mesh)
Red indicates significantly higher risk than that seen after anterior colporrhaphy (non-mesh)

4.8 Summary of findings for pelvic organ prolapse operations

Among the POP operations included in the analysis, open sacrocolpopexy and vaginal hysterectomy carried the highest risk of immediate complications. In general, infections and problems directly related to the operation were the most common immediate complications following prolapse operations. Excessive bleeding was also relatively common after open sacrocolpopexy and vaginal hysterectomy.

Mesh colporrhaphies (anterior and posterior) carried considerably higher risk of being readmitted for a complication over the five years following the initial operation than the equivalent operations carried out without mesh.

All procedures for vaginal vault prolapse (sacrospinous fixation, vaginal mesh vault repair, and open sacrocolpopexy) carried a higher risk of later complications than non-mesh anterior colporrhaphy; however complication rates were similar between the three different vault repair procedures.

Problems directly related to the operation, infections, and (for mesh operations) further surgery to remove the mesh were the most common later complications seen after operations to treat POP.

Mesh anterior colporrhaphy carried a higher risk of being readmitted for further incontinence surgery over the five years following the initial operation than non-mesh anterior colporrhaphy. Mesh colporrhaphies (anterior and posterior) also carried a higher risk of being readmitted for further prolapse surgery.

Procedures for vaginal vault prolapse (sacrospinous fixation, vaginal mesh vault repair, and open sacrocolpopexy) generally carried a higher risk of further incontinence and prolapse surgery than non-mesh anterior colporrhaphy; however further surgery rates were similar between the three different vault repair procedures.

4.9 How to interpret our findings

This study has used routinely available health information to look at:

• the number of operations provided in Scotland for SUI and POP; and
• how often women having the different types of operation develop problems after their surgery.

The study has several strengths. It includes all relevant operations provided in Scotland over a long time period. It uses high quality NHS information to assess how often women develop significant problems requiring further treatment up to five years following their initial operation.

The study also has some limitations. We only included first, single procedures so we cannot comment on outcomes following combined or repeat surgery. We only measured problems that required inpatient hospital treatment: problems that were dealt with in primary care or outpatient departments have not been included. We only described the number of problems seen so we cannot comment on the impact of these problems on patients' quality of life. In addition, it is possible that individuals with the most severe disease receive particular types of operation: this would then tend to make the results of those operations seem relatively poor.

4.10 Key Messages

• No operation is without risk. It is important for women and doctors to have clear information about the different risks associated with different types of operation. This will help them decide which operation will be best for an individual woman.
• The risk of immediate complications, later complications, and further surgery for SUI or POP differs between the different types of operation examined. An operation can carry a relatively high risk of one of these problems (for example immediate complications) but a relatively low risk of a different problem (for example longer term complications).
• More extensive operations, for example those involving operating through the abdomen or a hysterectomy, tend to carry the highest risk of immediate complications.
• Compared to open colposuspension, mesh tape operations for SUI tend to carry a lower risk of immediate complications, a lower risk of requiring subsequent prolapse surgery, and a similar risk of requiring repeat incontinence surgery. Mesh tape procedures carry a similar risk of longer term complications as open colposuspension (at least up to five years following the initial surgery), although the profile of later complications differs between the different operations, with subsequent mesh removal surgery only seen following mesh procedures. Our results currently support the use of mesh procedures for incontinence, although further research on longer term outcomes would be beneficial.
• Mesh colporrhaphies for the treatment of POP carry a substantially higher risk of later complications than non-mesh colporrhaphies. Mesh colporrhaphies also carry a higher risk of the woman needing further surgery for incontinence or prolapse than non-mesh colporrhaphies. Our results do not support the use of mesh colporrhaphies for primary prolapse repair.
• Procedures for vaginal vault prolapse repair in general carry higher risk than non-mesh anterior colporrhaphy, however rates of later complications and further incontinence and prolapse surgery are similar between vault repair procedures that do not involve mesh (sacrospinous fixation of the vagina) and those involving mesh inserted vaginally (vaginal mesh vault repair) or abdominally (open sacrocolpopexy). Our results do not clearly favour any particular vault repair procedure.